Developments and trends in enzyme catalysis in nonconventional media Sajja Hari Krishna* AK-Technische Chemie und Biotechnologie, Institut fu ¨r Chemie und Biochemie, Universita ¨t Greifswald, Soldmannstraße 16, D-17487 Greifswald, Germany Accepted 27 August 2002 Abstract The conventional notion that enzymes are only active in aqueous media has long been discarded, thanks to the numerous studies documenting enzyme activities in nonaqueous media, including pure organic solvents and supercritical fluids. Enzymatic reactions in nonaqueous solvents offer new possibilities for producing useful chemicals (emulsifiers, surfactants, wax esters, chiral drug molecules, biopolymers, peptides and proteins, modified fats and oils, structured lipids and flavor esters). The use of enzymes in both macro- and microaqueous systems has been investigated especially intensively in the last two decades. Although enzymes exhibit considerable activity in nonaqueous media, the activity is low compared to that in water. This observation has led to numerous studies to modify enzymes for specific purposes by various means including protein engineering. This review covers the historical developments, major technological advances and recent trends of enzyme catalysis in nonconventional media. A brief description of different classes of enzymes and their use in industry is provided with representative examples. Recent trends including use of novel solvent systems, role of water activity, stability issues, medium and biocatalyst engineering aspects have been discussed with examples. Special attention is given to protein engineering and directed evolution. D 2002 Elsevier Science Inc. All rights reserved. Keywords: Nonaqueous media; Enzyme catalysis; Medium engineering; Biocatalyst engineering; Protein engineering; Directed evolution 0734-9750/02/$ – see front matter D 2002 Elsevier Science Inc. All rights reserved. PII:S0734-9750(02)00019-8 * Tel.: +49-3834-864-366; fax: +49-3834-864-373. E-mail address: [email protected] (S. Hari Krishna). www.elsevier.com/locate/biotechadv Biotechnology Advances 20 (2002) 239 – 267
29
Embed
Developments and trends in enzyme catalysis in ...depa.fquim.unam.mx/amyd/archivero/CATALISIS-ENZIMATICA-INGENIE… · Developments and trends in enzyme catalysis in nonconventional
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Developments and trends in enzyme catalysis innonconventional media
Sajja Hari Krishna*
AK-Technische Chemie und Biotechnologie, Institut fur Chemie und Biochemie, Universitat Greifswald,Soldmannstraße 16, D-17487 Greifswald, Germany
Accepted 27 August 2002
Abstract
The conventional notion that enzymes are only active in aqueous media has long been discarded,thanks to the numerous studies documenting enzyme activities in nonaqueous media, including pureorganic solvents and supercritical fluids. Enzymatic reactions in nonaqueous solvents offer newpossibilities for producing useful chemicals (emulsifiers, surfactants, wax esters, chiral drugmolecules, biopolymers, peptides and proteins, modified fats and oils, structured lipids and flavoresters). The use of enzymes in both macro- and microaqueous systems has been investigated especiallyintensively in the last two decades. Although enzymes exhibit considerable activity in nonaqueousmedia, the activity is low compared to that in water. This observation has led to numerous studies tomodify enzymes for specific purposes by various means including protein engineering. This reviewcovers the historical developments, major technological advances and recent trends of enzymecatalysis in nonconventional media. A brief description of different classes of enzymes and their use inindustry is provided with representative examples. Recent trends including use of novel solventsystems, role of water activity, stability issues, medium and biocatalyst engineering aspects have beendiscussed with examples. Special attention is given to protein engineering and directed evolution.D 2002 Elsevier Science Inc. All rights reserved.
0734-9750/02/$ – see front matter D 2002 Elsevier Science Inc. All rights reserved.PII: S0734 -9750 (02 )00019 -8
* Tel.: +49-3834-864-366; fax: +49-3834-864-373.E-mail address: [email protected] (S. Hari Krishna).
www.elsevier.com/locate/biotechadv
Biotechnology Advances 20 (2002) 239–267
1. Introduction
According to conventional notion, enzymes are active only in water. Historically,enzymatic catalysis has been carried out primarily in aqueous systems. Although water is apoor solvent for preparative organic chemistry, it is the unique specificity of enzymes thatdrew the interest of chemists who were seeking highly selective catalytic agents.
Experiments to place enzymes in systems other than aqueous media date back to the end ofthe nineteenth century (Hill, 1898; Kastle and Loevenhart, 1900; Bourquelot and Bridel,1913; Sym, 1936; Dastoli and Price, 1967). Initial studies considered the addition of smallquantities of water-miscible organic solvents like ethanol or acetone to aqueous enzymesolutions ensuring availability of a high water content to retain the catalytic activity ofenzymes.
Then, biphasic mixtures (aqueous enzyme solution emulsified in a water-immisciblesolvent such as isooctane or heptane) were used, in which substrates from the organic phasediffuse to the aqueous phase, undergo enzymatic reaction and the products diffuse back. Thesize of water droplets may be reduced to facilitate mass transfer resulting in the formation ofmicroemulsions or reverse micelles, whose stabilization is achieved by adding surfactants(Martinek et al., 1986; Hari Krishna et al., 2002).
Developments in using enzymes in nearly nonaqueous solvents containing traces of waterhave stimulated research in achieving various kinds of enzymatic transformations (Klibanov,1986, 1989; Schoffers et al., 1996; Bornscheuer and Kazlauskas, 1999; Gandhi et al., 2000;Giri et al., 2001; Hari Krishna and Karanth, 2002a; Panke and Wubbolts, 2002; Thomas et al.,2002).
Enzymatic reactions in nonaqueous solvents offer numerous possibilities for the biotech-nological production of useful chemicals using reactions that are not feasible in aqueousmedia. These reactions include chiral synthesis or resolution (Klibanov, 1990; Collins et al.,1992; Stinson, 2000; Zaks, 2001); production of high-value pharmaceutical substances (Zaksand Dodds, 1997; Schulze et al., 1998; McCoy, 1999; Patel, 2001; Rasor and Voss, 2001);modification of fats and oils (Bornscheuer, 2000a); synthesis of flavor esters and foodadditives (Hari Krishna and Karanth, 2001, 2002b; Hari Krishna et al., 1999, 2000a,b,2001a,b); production of biodegradable polymers (Kobayashi, 1999), peptides, proteins andsugar-based polymers (Vulfson, 1998).
In nonaqueous solvents, hydrolytic enzymes can carry out synthetic reactions and someenzymes can also exhibit altered selectivities (Klibanov, 2001), pH memory (Zaks andKlibanov, 1985, 1988; Klibanov, 1995), increased activity and stability at elevated temper-atures (Zaks and Klibanov, 1984; Ahern and Klibanov, 1985), regio-, enantio- andstereoselectivity (Bornscheuer, 2000a,b), and may be affected by water activity (Halling,2000).
Currently, there is a considerable interest in the use of enzymes (particularly lipases, esterasesand proteinases) as catalysts in organic synthesis (Schmid and Verger, 1998; Bornscheuer,2000a,b; Carrea and Riva, 2000; Faber, 2000; Liese et al., 2000; Patel, 2000; Koeller andWong,2001). The diversity of potentially useful enzymes at the researcher’s disposal has now becomevast, supplemented by catalytic RNAs (ribozymes) and antibodies (abzymes).
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267240
Inorganic metal-derived catalysts are capable of carrying out several of the enzymaticreactions (Noyori, 1994), but the interest in using enzymes as agents for performing variousreactions with high selectivity in both macro- and microaqueous systems has picked uptremendously during the last two decades. On average, at least one paper dealing withbiocatalysis in organic solvents is published everyday. Since it would be an impossible task tocover every published account of biocatalysis in one review of reasonable length, this reviewfocuses on recent significant developments.
2. Enzymes as catalysts—a brief overview
Enzymes are classified into six major groups based on the reaction they catalyze (in the ECnumber order): oxido-reductases, transferases, hydrolases, lyases, isomerases and ligases. TheInternational Union of Biochemistry and Molecular Biology (IUBMB) recognizes almost4000 enzymes and has categorized them using enzyme nomenclature. The number of existingenzymes in nature is about 25,000. Latest information on enzymes is available on the Internet(http://www.expasy.ch/enzyme). Table 1 lists online accessible information resources onenzymes and proteins and various related aspects.
Enzymes occur widely not only in animals and plants but also in filamentous fungi,yeast and bacteria. Native or recombinant microorganisms produce a wide spectrum ofuseful enzymes with variations in substrate specificity, reaction rate, thermal stability andoptimum pH. Microbial enzymes are relatively easy to obtain by fermentation processesand with a few purification steps. Several enzymes from a variety of sources are availablefrom commercial suppliers (Table 2) and few vendors also offer enzyme-screening sets(Table 3).
Table 1Unified resource locators (URLs) for online accessible information on proteins and enzymes
Database URL
ASPDa http://www.sgi.sscc.ru/mgs/gnw/aspd/BRENDA http://www.brenda.uni-koeln.deENZYME http://www.expasy.ch/enzymeEnzyme Structures http://www.biochem.ucl.ac.uk/bsm/enzymesESTHER http://www.ensam.inra.fr/cholinesteraseKEGG http://www.genome.ad.jp/keggLIGAND http://www.genome.ad.jp/dbget/ligand.htmlLipase Engineering http://www.led.uni-stuttgart.deMDB http://metallo.scripps.eduMEROPS http://merops.iapc.bbsrc.ac.ukProtein Data Bank http://www.pdb.orgPROMISE http://bmbsgi11.leeds.ac.uk/promiseSWISS-PROT http://www.expasy.ch/sprotUM-BBD http://umbbd.ahc.umn.edu
a Artificially selected proteins and peptides database.
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 241
Of all the enzymes, hydrolases are the most employed for industrial biotransformations. Itis estimated that approximately 80% of all industrially used enzymes are hydrolases. Oxido-reductases are all cofactor-dependent and the industrial biotransformations have to beperformed coupled with processes for efficient recycling of the expensive cofactors. Trans-ferases and ligases, which play a far larger role in nature than in industry, will gain moreimportance in the near future. Lyases and isomerases are already gaining industrialsignificance for their unique properties. Special properties and limitations that are specific
Table 2Major commercial suppliers of enzymes
Supplier Website URL
Altus Biologics, Cambridge, MA, USA http://www.altus.comAmano Pharmaceutical, Nagoya, Japan http://www.amano-enzyme.co.jpAsahi Chemical Industry, Tokyo, Japan http://www.asahi-kasei.co.jpBiocatalysts, Pontypridd, UK http://www.biocatalysts.comBioCatalytics, Pasadena, CA, USA http://www.biocatalytics.comBiocon India, Bangalore, India http://www.biocon.comBiozyme Laboratories, South Wales, UK http://www.biozyme.comBoehringer Mannheim (now merged with Roche) http://www.roche.com/diagnosticsCalbiochem-Novabiochem, San Diego, CA, USA http://www.calbiochem.com
http://www.cnbi.comDiversa, San Diego, CA, USA http://www.diversa.com(Innovase LLC, a joint venture of Diversa with Dow Chemical) http://www.dow.com
DSM Food Specialties, Delft, The Netherlands http://www.dsm.nl/dfsFluka Chemical LLC, Buchs, Switzerland http://www.sigma-aldrich.comGenencor International, Rochester, NY, USA http://www.genencor.comGenzyme Biochemicals, UK http://www.genzyme.comGist-Brocades, The Netherlands (now DSM group) http://www.gist-brocades.nlHoechst, Germany (now Aventis, merged with Rhone-Poulenc) http://www.aventis.comJulich Enzyme Products, Julich, Germany http://www.juelich-enzyme.comLee Scientific, St. Louis, MO, USA http://www.leescientific.comMeito Sangyo, Tokyo, Japan http://www.mediagalaxy.co.jp/meitoMerck, Germany http://www.merck.comNagase and Co., Japan http://www.nagase.co.jpNew England Biolabs, Beverly, MA, USA http://www.neb.comNovozymes, Bagsvaerd, Denmark http://www.novozymes.comPromega, Madison, WI, USA http://www.promega.comRecordati, Milan, Italy http://www.recordati.itRoche Diagnostics, Mannheim, Germany http://indbio.roche.comRohm, Germany http://www.roehm.deSeppim, France http://www.sfrl.frServa, Germany (Invitrogen group) http://www.serva.deSigma-Aldrich-Fluka, St. Louis, MO, USA http://www.sigma-aldrich.comThermoGen, Woodridge, IL, USA http://www.thermogen.comToyobo, Tokyo, Japan http://www.toyobo.co.jpUnitica, Osaka, Japan http://www.unitica.co.jpWako Pure Chemicals Industries, Osaka, Japan http://search.wako-chem.co.jpWorthington Biochemical, Lakewood, NJ, USA http://www.worthington-biochem.com
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267242
for different enzyme classes with one representative of industrial example for each class areillustrated below.
2.1. Oxido-reductases (EC 1)
These enzymes catalyze oxido-reduction reactions, which means that they act on substratesthrough the transfer of electrons. The systematic name is based on donor/acceptor oxido-re-ductase. The oxidizing substrate is regarded as hydrogen donor and the enzyme is called adehydrogenase. If molecular oxygen (O2) is the acceptor, the enzymes may be named asoxidase.
All oxido-reductases are dependent on cofactors, which either supply or take the reducing oroxidizing equivalent. The most commonly needed cofactors are NADH/NAD + , NADPH/NADP + , FADH/FAD + , ATP/ADP and PQQ. Since most of them are expensive, an effectivecofactor regeneration system is required for a cost-effective industrial process. If an isolatedenzyme is being used, it is feasible to employ a second enzyme (in case of NADH, the bestapproach is to use formate dehydrogenase that utilizes formate and produces CO2, Scheme 1a)or by applying a second substrate (Scheme 1b). Application of whole cells instead of isolatedenzymes is also a viable choice. Although a fewmethods based on electrochemical regenerationhave been detailed (Ruppert et al., 1988), they have not yet developed to commercialization. Anexample of industrial processes employing benzoate dioxygenase (EC 1.14.12.10) is depictedin Scheme 2. Currently, these enzymes are gaining immense attention and their properties arebeing engineered mainly by directed evolution (Cirino and Arnold, 2002).
2.2. Transferases (EC 2)
These are enzymes that transfer a chemical group from one compound (donor) to another(acceptor). The systematic names follow the scheme donor/acceptor group transferase. Inmany cases, the donor is a cofactor or coenzyme carrying the often activated-chemical groupto be transferred.
These play a larger role in nature than in industry. Very few transferases are used inindustry due to various reasons: equilibrium reactions often do not attain high yields,coupling reactions occur and the group-transferring substrates are quite expensive or theircorresponding products are not easily recycled. However, these enzymes would gainimportance in the future if the problems associated with them can be solved. Nonetheless,
Table 3Commercially available enzyme screening kits
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 243
high regio- and stereoselectivities in transferase-catalyzed reactions are major causes for theirincreasing utility. An industrial process employing D-amino acid transaminase (EC 2.6.1.21)is presented in Scheme 3.
Scheme 2. Enzymatic transformation catalyzed by an oxido-reductase.
Scheme 1. Different approaches of cofactor regeneration.
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267244
2.3. Hydrolases (EC 3)
These enzymes catalyze the hydrolytic cleavage of C)O, C)N, C)C and some other bonds,including P)O bonds in phosphates. Their applications are very diverse including hydrolysisof polysaccharides, nitriles, proteins, lipids, and esterification of fatty acids. Most of theseenzymes are used in processing-type reactions to degrade proteins, carbohydrates and lipids, indetergent formulations, and in the food industry (Table 4). The term hydrolase is included inthe systematic name, which includes the name of the substrate and the suffix –ase.
Hydrolases are the most commonly used enzymes in industrial processes. Biomassutilization, with the help of cellulases (a group of hydrolases) to produce fine chemicals(Hari Krishna et al., 2001c), is also a major thrust area of research in various laboratories. Theliterature pertaining to hydrolases is well documented in numerous reviews (Theil, 1995;Schmid and Verger, 1998; Rao et al., 1998; Gandhi et al., 2000; Koeller and Wong, 2001;Sharma et al., 2001; Bornscheuer, 2002; Gupta et al., 2002; Hari Krishna and Karanth, 2002a)and books (Koskinen and Klibanov, 1996; Bornscheuer, 2000a; Bornscheuer and Kazlauskas,1999; Faber, 2000; Liese et al., 2000; Patel, 2000; Drauz and Waldmann, 2002). Arepresentative example of industrial processes employing porcine pancreatic lipase (EC3.1.1.3) is showed in Scheme 4.
2.4. Lyases (EC 4)
These enzymes catalyze the cleavage of C)C, C)O, C)N and a few other bonds in adifferent fashion from hydrolysis, often leaving double bonds that may be subjected to furtherreactions. Systematic denomination follows the pattern substrate group-lyase. The hyphen isimportant to avoid any confusion, e.g., the term hydro-lyase should be used instead ofhydrolyase, which looks quite similar as hydrolase.
Scheme 3. Enzymatic transformation catalyzed by a transferase.
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 245
Table 4Some of the important industrial processes using hydrolasesa
Industry Products (application) Enzyme and process Medium Remarkb
BASF Chiral amines andalcohols (intermediatesfor pharmaceuticalsand pesticides, chiralsynthons in asymmetricsynthesis)
Burkholderia plantarii
lipase (immobilized);hydrolysis
MTBE-ethylmethoxyacetate
E> 500(>100 ton)
Bristol-Myers-Squibb(BMS)
(3R,4S)-Azetidinoneacetate (used in thesynthesis of paclitaxel,Taxol)
P. cepacia lipase(immobilized); hydrolysis
Aqueous ee>99%(multi kg)
Bristol-Myers-Squibb Hydroxy methyl glutarylcoenzyme A (HMG-CoA)reductase inhibitor(anticholesterol drug)
P. cepacia lipase(immobilized); acetylation
Toluene ee = 98%(multi kg)
Chiroscience Intermediate for theanti-HIV agent carbovir,hypocholesteremicreagents and antifungalagent brefeldin A
P. fluorescens lipase(soluble); hydrolysis
Aqueous ee>92%(multi kg)
DSM; Tanabe Seiyaku Generation of intermediatefor synthesis of Diltiazem(antihypertensive drug)
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267246
Lyases have gained significant industrial attention as the predominant and natural bond-breaking (lyase) reactions can be reversed (bond formation, i.e., lyase acting as synthetase)under nonnatural conditions (i.e., high reactant concentrations), leading to the construction ofnew bonds of commercial importance. Often chiral centers are generated during bondformation. Using protein engineering techniques, the substrate range of these enzymes iscurrently being expanded. An industrial example employing phenylalanine ammonia-lyase(EC 4.1.99.2) is presented in Scheme 5.
2.5. Isomerases (EC 5)
This class represents a small number of enzymes that catalyze geometric or structuralchanges within one single molecule and make it possible to employ cheaper substrates to
Industry Products (application) Enzyme and process Medium Remarkb
Coca-Cola (S)-Phenylalanine(intermediate in thesynthesis of aspartame)
a Data from Liese et al. (2000); Bornscheuer (2000b).b E is the E-value (enantioselectivity), ee is enantiomeric excess.
Table 4 (continued)
Scheme 4. Enzymatic transformation catalyzed by a hydrolase.
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 247
obtain high-value products. Depending on the type of isomerism, these enzymes may betermed as epimerases, racemases, cis–trans isomerases, tautomerases or mutases. Racemasesare particularly important in kinetic resolutions. The most renowned candidate of this groupis certainly glucose isomerase (EC 5.3.1.5), whose industrial application is depicted inScheme 6.
2.6. Ligases (EC 6)
These catalyze a bond formation between two molecules, coupled with hydrolysis of apyrophosphate bond in ATP or similar triphosphate. The bonds formed are C)O, C)S andC)N. The systematic names are written as X:Y ligase. No industrial process is carried outusing ligases at a kilogram scale, but they play a significant role in nature (in ribosomalpeptide synthesis) and also in repairing DNA fragments and in genetic engineering (e.g.,DNA ligases catalyze C)O bond formation in DNA synthesis).
Extensive information on the potential of different classes of enzymes in organic synthesisis provided in various books (Koskinen and Klibanov, 1996; Faber, 2000; Liese et al., 2000;Drauz and Waldmann, 2002).
Scheme 5. Enzymatic transformation catalyzed by a lyase.
Scheme 6. Enzymatic transformation catalyzed by an isomerase.
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267248
3. Early developments
3.1. A fascinating history
Enzymes have been in use for thousands of years before their nature was actuallyunderstood. No one really knows when calf stomach was first used as a catalyst for thefirst time in the manufacture of cheese. As early as 1783, Spallanzani demonstrated thatgastric juice could digest meat in vitro (Schwan called the active substance responsible forthis as pepsin in 1836 and Berzelius described a ‘catalyst’ as a substance which can breathelife into slumbering chemical reactions in 1835) (Gutfreund, 1976).
Kirchhoff, in 1814, observed sugar production from starch by malted barley. The activeprinciple of malt was called diastase and its application was described by Payen and Persoz in1833. Dubonfout observed invertase activity in 1846. Further to these, Kuhne suggested, in1876, that such nonorganized ferments should be called ‘‘enzymes.’’ The terms ‘‘organizedferment’’ (e.g., cell-free yeast extract) and ‘‘unorganized ferment’’ (e.g., gastric juice secretedby cells) are no longer used. Kuhne also presented some interesting results from his studieswith trypsin. In 1893, Ostwald defined ‘‘catalyst’’ and classified enzymes as catalysts.
In 1894, Emil Fischer observed that the enzyme called ‘emulsin’ catalyzes the hydrolysisof b-methyl-D-glucoside, while ‘maltase’ is active towards a-methyl-D-glucoside and formu-lated his ‘‘lock-and-key’’ theory of enzyme specificity. In 1897, Buchner demonstrated theconversion of glucose to ethanol using cell-free extract from yeast. Warburg carried outpreparative separation of L-leucine from a racemic mixture by hydrolyzing the propyl esterwith liver extracts in 1906. Rosenberg used D-oxynirerilase from almonds as catalyst for thesynthesis of optically active cyanohydrins in 1908.
In the area of nonaqueous biocatalysis, Hill (1898) was the first to observe that thebiocatalysis of hydrolytic enzymes is reversible. Pottevin (1906) demonstrated that crudepancreatic lipase could synthesize methyl oleate from methanol and oleic acid in a largelyorganic reaction mixture. Bourquelot, Bridel and Verdon described glucoside synthesis in thepresence of high concentration of ethanol or acetone between 1911 and 1913 (Bourquelot andBridel, 1913). Sym (1936) improved the enzymatic ester synthesis using pancreatic lipase inpresence of benzene.
Michaelis and Menton (1913) published a theoretical consideration of kinetics ofenzymatic catalysis, which led to the development of the so-called M–M equation. In1916, Nelson and Griffin demonstrated immobilization of invertase on charcoal with activityretention. By 1920, about a dozen enzymes were known, but none of them had been isolated.In 1926, James Sumner crystallized urease from jack bean (Canavalia ensiformis) anddemonstrated that it is a simple protein, which was confirmed by Northrop. By the late 1940s,many enzymes were available in pure form, but there was still no evidence that proteinspossessed unique amino acid sequences. In 1953, Sanger established the first primarysequence of a protein (insulin), proving the chemical identity of proteins. Lysozyme wasthe first enzyme whose 3-D structure was defined in 1967 with the help of X-raycrystallography (Phillips, 1967). Further to this, Gutte and Merrifield (1969) synthesizedwhole sequence of ribonuclease A in 11,931 steps on a laboratory scale by organic chemical
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 249
methods. Currently, about 4000 enzymes have been catalogued and several hundred of thesecan be obtained commercially.
While the first benefit for the industry from the microbiological development had comevery early, enzyme catalysis hardly influenced the industry at that time (Moo-Young et al.,1985; Davies et al., 1989; Roberts et al., 1995). The commercial use of extracellular microbialenzymes started in the West around 1890, stimulated by a Japanese entrepreneur namedTakamine, who produced enzymes in the USA based on Japanese technology. The principalproduct was ‘Taka-diastase’, which was a mixture of amylolytic and proteolytic enzymesobtained from Aspergillus oryzae. Rohm (Darmstadt, Germany) applied pancreatic enzymesin the leather processing for bating of hides in 1908, and in the cleaning of laundry between1913 and 1915. In France, Boidin and Effront developed bacterial enzymes in 1913 andfound that Bacillus subtilis produced a thermostable a-amylase when grown in still cultureson a liquid medium prepared by extraction of malt or grain. In 1960, Novo Nordisk(Bagsvaerd, Denmark) produced protease on a large-scale by cultivating B. licheniformisin submerged culture.
Invertase was probably the first immobilized enzyme to be used commercially for theproduction of Golden Syrup by Tate and Lyle (Decatur, IL, USA) during World War II,because the preferred reagent, sulfuric acid, was unavailable at that time (Cheetham,1995). Industrial processes for L-amino acid production based on the batch use ofsoluble aminoacylase were already in use in 1954. However, like many batch processeswith soluble enzymes, they had their disadvantages such as high labor costs, compli-cated product separation, low yields, high enzyme costs and difficulty in recycling theenzyme.
During the mid-1960s, Tanabe Seiyaku (Tokyo, Japan) tried to overcome these problemsby using immobilized aminoacylases, and eventually produced L-methionine, in 1969, byaminoacylase immobilized on DEAE-Sephadex in a packed-bed reactor. This became the firstlarge-scale use of an immobilized enzyme (Trevan, 1980). In 1980, Degussa (Dusseldorf,Germany) developed a membrane reactor system with native enzymes in homogeneoussolution for the large-scale production of enantiomerically pure L-amino acids (Bommarius etal., 1992).
Enzymatic isomerization of glucose to fructose represents the largest use of immobilizedenzyme in the manufacture of fine chemicals. High-fructose corn syrup has grown to becomea large-volume biotransformation product. While sucrose is sweet, fructose is about 1.5 timessweeter and consequently in high demand as a sweetener. However, the food industry took along time to become acquainted with the potential of glucose isomerase. The Japanese werethe first to employ soluble glucose isomerase to produce high-quality fructose syrups in 1966.In 1967, Clinton Corn Processing (Clinton, IA, USA) manufactured enzymatically producedfructose corn syrup and started, in 1974, the commercial production of fructose syrups usingglucose isomerase immobilized on a cellulose ion-exchange polymers. In 1976, Kato Kagaku(Kohwa, Nagoya, Japan) was first to manufacture these syrups in a continuous process asopposed to a batch process and in 1984, to isolate crystalline fructose. The glucose isomerase,Sweetzyme T (which has a long life because of immobilization), produced by Novo Nordiskis now largely used in the starch processing industry. Central del Latte (Milan, Italy) was the
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267250
first company to commercially hydrolyze milk lactose using SNAMprogetti technology withimmobilized lactase.
Penicillin G, discovered accidentally in 1929 by Fleming in Penicillium notatum,revolutionized the practice of medicine. Currently, thousands of semisynthetic b-lactamantibiotics are in production. Most of these are prepared from 6-aminopenicillanic acid (6-APA), 7-aminocephalosporanic acid (7-ACA) and 7-amino-des-acetoxycephalosporanic acid(7-ADCA). Presently, 6-APA is mainly produced by chemical or enzymatic (using penicillinamidase) deacylation of penicillin G or penicillin V. This process, used since 1973, is thebest-known use of an immobilized enzyme in the pharmaceutical industry. In 1979,enzymatic production of 7-ACA was realized when Toyo Jozo (now acquired by AsahiKasei, Tokyo, Japan), in collaboration with Asahi Chemical Industry (Tokyo, Japan),developed a chemo-enzymatic two-step process starting from cephalosporin C (Scheme 7).About 90 tons/annum of 7-ACA are produced using this technology.
Since the 1980s, many industries and research laboratories are applying genetic engin-eering techniques to improve enzyme production and to alter enzyme properties throughprotein engineering and evolutionary design.
3.2. Major technological advances
Five major technological advances are believed to have significantly influenced industry inadopting enzymatic biotransformations (Lilly, 1994): (i) the development of large-scaledownstream processing techniques for the release of intracellular enzymes from the micro-organisms; (ii) improved screening methods for novel biocatalysts (Kieslich et al., 1998;Demirjan et al., 1999; Miller, 2000; Asano, 2002; Ornstein, 2002); (iii) the development ofimmobilized enzymes; (iv) biocatalysis in organic media; and most recently (v) recombinant-DNA (r-DNA) technology to produce enzymes at a reasonable cost.
Buckland et al. (1975) examined the use of high proportions of organic solvents to increasethe solubility of reactants in production of cholestenone using isolated cholesterol oxidase.There seems to be no agreement as to why the biocatalysis in organic media did not take offearlier (Halling and Kvittingen, 1999; Klibanov, 2000; Kvittingen, 2000). Perhaps thetraditional belief that most enzymes are incompatible with most organic syntheses innonaqueous media posed a psychological hurdle. Also, until recently, there was no demandfor enantiopure compounds, and hence, no need to use enzymes. The establishment ofindustrial processes (Coleman and Macrae, 1977; Matsuo et al., 1981), and the realization thatmost enzymes can function well in organic solvents (Zaks and Klibanov, 1984, 1985, 1986,1988), have heightened interest in the use of enzymes. Also, the need for enantiomericallypure drugs is driving the demand for enzymatic processes. This combined with the discoveryof striking new properties of enzymes in organic solvents has led to establishment of organicphase enzyme processes in industry (Bornscheuer, 2000b; Liese et al., 2000).
Generally, microorganisms isolated from nature produce the desired enzymes at levels thatare too low to allow a cost-effective process. Consequently, a modification of the organism isdesirable for enhancing enzyme production. For this, three approaches are followed for strainimprovement at present. The first one—directed evolution—involves improvement by
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 251
mutation of the gene concerned and selection or screening (Arnold and Moore, 1997;Bornscheuer et al., 2002). The second method is hybridization, which involves modificationof the cellular genetic information by transferring of DNA from another strain. The thirdmethod is r-DNA technology, whereby genetic information from one strain is manipulated invitro and then inserted into the same or another strain.
Scheme 7. Chemo-enzymatic two-step process for the production of 7-amino cephalosporanic acid (7-ACA) fromcephalosporin C.
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267252
Random mutagenesis using UV light or chemical mutagens is a well-established generaltool for strain improvement, but the excitement of directed evolution stems from the ability totarget a single gene or region of a gene for achieving the desired effect. Directed evolution(also called molecular evolution or in vitro evolution) involves using PCR, expressing theprotein and then selecting or screening for strains with improved properties. The DNAshuffling technique (Stemmer, 1994a,b; Stemmer et al., 1995) places the directed evolutionapproach apart from earlier random mutagenesis and screening efforts.
The r-DNA technology has dramatically changed the enzyme production scene. Proteinengineering combined with recombinant expression systems allows one to produce a newenzyme variant and to rapidly attain concentrations that make industrial production worth-while. Many microbial enzyme genes have been cloned over the past few years, includinggenes for important commercial enzymes. High-level expression has been achieved mainly inPichia pastoris, Saccharomyces cerevisiae and Escherichia coli. The secretion is facilitatedby a leader sequence fused to the expressed gene. As P. pastoris does not secrete significantamounts of proteins into the medium naturally, the a-factor prepro-peptide from S. cerevisiaeis used to induce secretion. Genes will be under the control of strong methanol-inducible AOXpromoter (in case of P. pastoris) or galactose-inducible GAL10 promoter (in case of S.cerevisiae). In contrast to S. cerevisiae, P. pastoris does not hyperglycosylate the proteins andcan provide up to 100-fold higher expression levels. For the expression of bacterial enzymes,the genes are placed under the control of the strong temperature-inducible l phage promoterPL or rhamnose-inducible promoter on the E. coli expression vector. The choice of anappropriate expression system combined with optimal expression with genetic modificationsas well as suitable fermentation conditions allows the effective production of large amounts ofrecombinant enzymes.
4. Recent trends
Enzymes occupy a unique position in synthetic chemistry due to their high selectivities andrapid catalysis under ambient reaction conditions. Nevertheless, synthetic chemists have beenreluctant to employ enzymes as catalysts, because most organic compounds are water-insoluble, and the water removal is tedious and expensive. The realization that enzymes canretain and, in some cases, improve their high specificity in nearly anhydrous media, hasdramatically changed the prospects of employing enzymes in synthetic organic chemistry.The problems that arise for most biotransformations are low solubility of reactants andproducts, and limited stability of biocatalysts.
Carrying out reactions in an aqueous-organic two-phase system would be a solution toovercome the first problem. This is not always possible due to the limited stability of enzymesat liquid–liquid interface or in organic solvents. Hence, other approaches are necessary.These include addition of complexing agents such as dimethylated cyclodextrins or adsorbingmaterials like XAD-7 resins (Eli Lilly, Indianapolis, USA), use of membrane-stabilizedinterface (Sepracor, Marlborough, MA, USA) and continuous extraction of reaction products(Forschungszentrum-Julich, Julich, Germany).
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 253
The catalyst’s stability can be increased using a variety of methods including the addition ofantioxidants (e.g., dithiothreitol), immobilization, cross-linking, separation from deactivatingreagents, variation of reaction conditions, and by genetic engineering (Burton et al., 2002).
4.1. Solvent systems
Solvent systems used as the reaction media for enzymatic catalysis may be categorized as:(1) aqueous; (2) water: water-miscible (monophasic aqueous-organic system); (3) water:water-immiscible (biphasic aqueous-organic system); (4) nonaqueous (monophasic organicsystem); (5) anhydrous; (6) supercritical fluids; (7) reversed micelles; (8) solvent-freesystems; (9) gas phase; and (10) ionic liquids. More information can be found in the reviewsand books recommended above. A recent development is the use of ionic liquids for enzymecatalysis (Madeira-Lau et al., 2000), to improve activity, stability and selectivity (Park andKazlauskas, 2002). The most common ionic liquids employed are tetrafluoroborate andhexafluorophosphate.
4.2. Importance of water activity
Water is critical for enzymes and affects enzyme action in various ways: by influencingenzyme structure via noncovalent bonding and disruption of hydrogen bonds; by facilitatingreagent diffusion; and by influencing the reaction equilibrium. Too low a water contentgenerally reduces enzyme activity. A high water content can also reduce reaction rates byaggregating enzyme particles and causing diffusional limitations. The optimum amount ofwater is often within a narrow range. To quantify the amount of water present in the reactionmixture, the thermodynamic water activity (aw) is the preferred measure.
Optimal water activity is not only important to maintain the catalytic activity of an enzyme,but also to obtain high reaction rates and yields, and stability of the biocatalyst. Completelyanhydrous solvents do not support enzymatic activity. Some water is always necessary for theenzyme to retain its native structure responsible for catalysis. The amount of water required toretain catalytic activity is enzyme dependent. a-Chymotrypsin needs only 50 molecules ofwater per enzyme molecule to remain catalytically active (Zaks and Klibanov, 1986).
Enzymes like subtilisin and various lipases are similar in their requirement for tracequantities of water (Zaks and Klibanov, 1988), while much more water is required for someenzymes (e.g., polyphenol oxidase requires the presence of about 3.5! 107 molecules ofwater) (Zaks and Klibanov, 1985). The water present in a biological system can be separatedinto two physically distinct categories: the majority (>98%) serving as a true solvent (bulkwater) and a small portion being tightly bound to the enzyme’s surface (bound water). Boundwater is crucial for the enzyme structure and activity.
4.3. Stability aspects
Enzymes inactivate at high temperatures in aqueous media due to both partial unfoldingand covalent alterations in the primary structure (Ahern and Klibanov, 1985). Water is
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267254
required for both these mechanisms and hence, the enzyme thermostability in nonaqueousenvironments should be high mainly as a consequence of protein rigidity in these systems.For instance, porcine pancreatic lipase remained stable at 100 !C for >12 h in nonaqueoussolvents (Zaks and Klibanov, 1984). Such stability is seldom possible with other approacheslike chemical cross-linking, immobilization or protein engineering.
Another reason for thermostability, apart from rigidity, is that a number of covalentprocesses involved in irreversible or reversible inactivation of proteins such as deamidation,peptide hydrolysis and cystein decomposition require water, but are extremely slow innonaqueous systems. Most of the early work had been restricted to relatively nonpolarsubstrates. Stability studies involving polar and water-soluble substrates are also essential tobring out a clear picture. Presently, enzymatic catalysis is also carried out in gas phase(Lamare and Legoy, 1993), supercritical fluids (Kamat et al., 1995) and ionic liquids (Parkand Kazlauskas, 2002), for which enhanced thermostability of enzymes is important.
Despite a higher thermostability, the reaction rates of enzyme in nonaqueous systems arelow compared with the reaction rates in aqueous media (Klibanov, 1997). Thermostabiliza-tion of enzymes also results sometimes in stabilization towards other denaturing conditions.Arnold (1990) suggested correlations between enhanced thermostability and stability innonaqueous systems. In this respect, proteins from extremophiles do not differ significantlyfrom their mesophilic counterparts (Haney et al., 1999; Eichler, 2001; Cavicchioli et al.,2002). However, no general rule or strategy of stabilization has yet been framed, althoughseveral attempts have been made to generalize (Fagain, 1995; Russell and Taylor, 1995;Shoichet et al., 1995; Hendsch et al., 1996; Jaenicke, 2000; Jaenicke and Bohm, 1998; Kumaret al., 2000; Lehmann and Wyss, 2001; Lehmann et al., 2002).
During synthetic reactions (e.g., esterification), complications arise because water is aproduct. It is observed that, in addition to decreasing the equilibrium yield, accumulation ofwater results in decreased enzyme activity. Moreover, accumulated water has also beenshown to adversely affect the long-term stability of the enzyme (Hari Krishna et al., 2001a).The reduction in water content of the enzymes usually results in an increase of their half-lives.
4.4. Enzyme stabilization strategies
The approaches followed to stabilize enzymes are mainly two: (i) medium engineering and(ii) biocatalyst engineering.
4.4.1. Medium engineeringEnzymes in nonaqueous systems can be active provided that the essential water layer
around them is not stripped off. Medium engineering in the context of biocatalysis innonaqueous media involves the modification of the immediate vicinity of the biocatalyst.Nonpolar solvents are better than polar ones since the former provide a better microenviron-ment for the enzyme. If the enzyme’s microenvironment favors high substrate and lowproduct solubility, the reaction rates would be high.
The solvent effects may not be generalized too far. There are various exceptions of whichlipases are a particular case. For instance, porcine pancreatic lipase is active in anhydrous
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 255
pyridine (Zaks and Klibanov, 1985), suggesting that it can retain the bound water even in awater-miscible solvent. Since the natural environment for lipases is nonpolar, the ability tobind essential water tightly might have evolved as a prerequisite condition for catalysis inhydrophobic environments.
Several aspects have to be considered in choosing an appropriate solvent for a givenreaction. These include compatibility with the selected reaction (substrates and products),inertness, a low density to minimize mass transfer limitations, and other properties that aresuitable (e.g., surface tension, toxicity, flammability, waste disposal and cost). Halling (1994,2000) presented a detailed account of predictions that can be made to elucidate the influenceof solvent selection on the equilibrium. Reetz (2002a) reviewed various medium engineeringsuccesses with particular reference to lipase catalysis.
4.4.2. Biocatalyst engineeringImmobilization and protein engineering are long known for improving biocatalyst
efficiency or stability of enzymes in aqueous media. Also, these methods have been usedto improve biocatalyst performance in organic solvents. Laane (1987) used the term‘‘biocatalyst engineering’’ for these approaches.
Various developments have taken place in this area including the generation of active andstable homogeneous (soluble) biocatalysts by the covalent or noncovalent modification of thenative enzyme. Covalent techniques are well described in the literature (e.g., attachment ofpolyethylene glycol chains to enzymes), while noncovalent modifications are less common,although they are able to provide highly active and soluble enzyme forms (Okahata and Mori,1997). These developments have been well documented in various reviews (Khmelnitsky etal., 1988; DeSantis and Jones, 1999a,b; Govardhan, 1999; Villeneuve et al., 2000).
Enzymes have been employed in nonaqueous systems in various states such as nativeenzymes, suspended enzyme powder, solid enzyme adsorbed on support, polyethyleneglycol-modified enzymes soluble in organic solvents (Inada et al., 1986), enzyme entrappedwithin a gel or microemulsion or reversed micelle (Bornscheuer et al., 1999a) andimmobilized enzyme. No general guidelines are available yet for choosing the mostappropriate form of the enzyme for a specific purpose.
Enzymes are frequently prepared from aqueous/buffer solution via lyophilization, whichresults in undesirable changes in the protein’s secondary structure and about 40% of the activesites may be denatured. The addition of specific small molecules (excipients) in the freeze-drying stage often improves catalytic activity. This formed the basis of molecular imprinting,which involves the complex formation between a macromolecule and low-molecular-weightligand in solution, followed by drying and washing with a selective solvent that removes theligand. The protein retains the ligand-induced conformation even after the removal of theligand (Russell and Klibanov, 1988).
Activation has been achieved by the addition of crown ethers (Engbersen et al., 1996; vanUnen et al., 2002), lyoprotectants (Dabulis and Klibanov, 1992), transition-state analogues(Slade and Vulfson, 1998) and substrate or substrate mimics or competitive inhibitors(Russell and Klibanov, 1988). Nonligand lyoprotectants (sorbitol, sugars and PEG) alsoenhanced enzyme activity in organic solvents when present during lyophilization. However,
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267256
addition of excipients to suspensions of native enzymes in organic solvents had noappreciable effect, indicating that the interaction of the excipient with soluble enzyme isessential to alleviate denaturation of enzymes during lyophilization (Klibanov, 2001).Addition of surfactants or hydrophobic sol–gel materials before lyophilization enhancedthe lipase activity in organic solvents by up to 100-fold (Reetz, 2002b).
Addition of an inorganic salt has been shown to dramatically enhance the activity ofenzymes in organic solvents (Bedell et al., 1998). Ion pairing of biocatalysts in the presenceof very low concentrations of ionic surfactants resulted in remarkably active ion-pairedenzymes (e.g., subtilisin and a-chymotrypsin with >1000-fold higher activities over nativeenzymes were generated). Ion-paired enzymes have also been incorporated into plasticmaterials for preparing ‘‘biocatalytic plastics’’ (Wang et al., 1997).
Activated biocatalysts have already found application in the pharmaceutical industry. Salt-activated thermolysin (a bacterial protease) is used to selectively acrylate the 20-hydroxylgroup of taxol in tert-amyl alcohol. In a specific case, taxol was acylated with divinyladipateto yield taxol 20-vinyladipate, which was used as an acyl donor in Candida antarctica lipase-catalyzed hydrolysis of the terminal vinyl ester to get taxol 20-adipic acid derivative that is" 1700 times more water-soluble and can be used to design taxol prodrugs with increasedbioavailability (Schmid et al., 2001).
Roche (Mannheim, Germany) and Novo Nordisk market several adsorption-immobilizedenzymes. Altus Biologics (Cambridge, MA, USA) offers various cross-linked enzymecrystals (CLECs) of different enzymes. Fluka (Buchs, Switzerland) supplies sol–gelentrapped enzymes.
4.5. Protein engineering
Despite hard competition from chemical modification and related techniques, proteinengineering has become an increasingly important strategy for improving enzymes (Rubingh,1997; Cedrone et al., 2000; Chen, 2001). The importance of protein engineering in industrycontinues to grow with the expanding range of protein applications. Extremophilic proteinsisolated from organisms from extreme environments are emerging as an important source ofnew backbones for engineering proteins to attain new properties. The current strategies ofprotein engineering include rational design and directed evolution.
4.5.1. Rational designThis was the earliest approach to protein engineering and is still widely employed to
introduce desired characteristics into a target protein. The growing understanding of how toengineer certain basic enzyme properties (e.g., stability, activity and surface properties) isbeginning to make rational design more efficient. Advances in rational design depend on theprogress made in structure determination, improved modeling protocols and significant newinsights into structure–function relationships (Fersht, 1999). Advances in modeling of freeenergy perturbation methods and molecular dynamics calculations also influence this area.Moult (1996) discussed the state-of-the-art in comparative modeling and ab initio proteinstructure predictions.
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 257
Two impressive cases of rational design were reported recently. In the first example,enantioselectivity of C. antarctica lipase-catalyzed resolution of 1-chloro-2-octanol wasimproved (E= 14 to 28) by a single amino acid exchange, Ser-47–Ala, as predicted bymolecular modeling (Rotticci et al., 2001). In the second case (Magnusson et al., 2001), atransition-state stabilizing threonine near the active site was removed by site-directedmutagenesis and the lost activity was restored by using chiral substrate (2-hydroxy-propanoate) containing the missing functional –OH group, which resulted in improvedenantioselectivity (E = 1.6 to 22). Creating tailored enzymes with opposite enantiopreferencemay also be possible with this approach.
Although use of modeling to predict the enantioselectivity is being used by numerousresearch groups, much of the available data is empirical and is not easily interpreted at themolecular level (Kazlauskas, 2000). This is largely due to ambiguity in identifying theparameters responsible for enzyme–substrate interactions. In addition, properties like activity,specificity and stability are controlled by multiple sites on protein. Substantial progress isessential in establishing these features, which would aid application of rational design in atruly directed manner (Ottosson et al., 2002). A special issue of Biochimica et BiophysicaActa: Protein Science and Molecular Enzymology has been devoted to protein engineeringspanning the advances achieved with various enzymes (Dalboge and Borchert, 2000),indicating the importance of this exciting area of research.
4.5.2. Directed evolutionWhile site-directed mutagenesis and rational protein design are widely practiced, an
alternate method—directed evolution—is gaining increased attention from academic andindustrial laboratories to modify and improve important biocatalysts (Stemmer et al., 1995;Chirumamilla et al., 2001) and to achieve various other objectives (Patten et al., 1997;Chartrain et al., 2000). This strategy combines random mutagenesis of target gene withscreening or selection for the desired property and is especially useful for cases like solventtolerance or thermostability where available theories are inadequate to make predictions. Infact, earlier protein engineering studies were targeted to adapt proteins for nonnaturalenvironments (Arnold, 1988, 1990, 1993a,b; Chen and Arnold, 1991, 1993; Chen et al.,1991).
The goals currently envisaged are to improve the enzyme’s activity, stability andselectivity. Much of current research is directed towards increasing the enzyme stability.Rational design, in this context, is inferior because the molecular basis for increased stabilityis ill-defined. Thermostability is also difficult to improve rationally and hence, is a goodtarget for directed evolution. Random mutagenesis and screening have resulted in morethermostable subtilisin and lipase (Shinkai et al., 1996). Contrary to these successes, arational approach to increase the thermostability of P. camembertii lipase by introducing adisulfide link was a failure (Yamaguchi et al., 1996).
Numerous choices are available for creating DNA libraries: e.g., error-prone PCR,combinatorial oligonucleotide mutagenesis, DNA shuffling, exon shuffling, random-primingrecombination, random chimeragenesis on transient templates (RACHITT), staggered exten-sion process (StEP recombination), heteroduplex recombination, incremental truncation for
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267258
Table 5Selected examples for the directed evolution of biocatalysts
Biocatalyst Evolution target (result) Reference
P. aeruginosa lipase Increased enantioselectivity (E = 1 to 26) Reetz et al. (1997)Staphylococcus hyicus
and S. aureus lipasesAltered substrate specificity,phospholipids versus short-chainfatty esters (up to 12-fold increase)
van Kampen andEgmond (2000)
P. fluorescens esterase Increased selectivity (active mutants obtained) Bornscheuer et al. (1999b),Henke and Bornscheuer (1999)
B. subtilis p-nitrobenzylesterase
Increased thermostability (14 !C increase inTm without decrease in original activity atlow temperatures)
Giver et al. (1998)
p-Nitrobenzyl esterase Activity in organic solvents (50–150-foldactivity in 25–30% DMF)
Moore and Arnold (1996)
B. sphaericus protease Increased activity at 10 !C (6-fold increase) Wintrode et al. (2000)B. lentus subtilisin Expression level of secreted enzyme (50% increase) Naki et al. (1998)Various subtilisins Overall improvement of various properties
(active chimaeras with increased activityand stability)
Ness et al. (1999)
Subtilisin BPN0 Increased activity at low temperatures(2-fold increase in rate at 10 !C)
Taguchi et al. (1999)
Subtilisin E Enhanced thermostability(significant increases reported)
Zhao and Arnold (1999),Zhao et al. (1998),Miyazaki and Arnold (1999)
b-Lactamase Increased activity toward cefotaxime(32,000-fold and 2383-fold increase)
Stemmer (1994a);Zaccolo and Gherardi (1999)
Cephalosporinase Increased activity toward moxalactam(up to 540-fold increase)
Crameri et al. (1998)
Carboxymethylcellulase
Increased activity(up to 5-fold increase)
Kim et al. (2000)
Galactosidase Activity switched to fucosidase(66-fold increase in specific activity)
Zhang et al. (1997)
Cytochrome P-450 Increased activity toward naphthalenewithout cofactors (up to 20-fold)
Joo et al. (1999)
Peroxidase (fungal) Increased temperature and oxidativestability (>100-fold in both targets)
Cherry et al. (1999)
Catalase I Increased peroxidase property(2% to 58% increase)
Matsuura et al. (1998)
3-Isopropylmalatedehydrogenase
Increased thermostability(>3-fold activity at 70 !C)
Akanuma et al. (1998)
Lactate dehydrogenase Increased activity without cofactor(70-fold increase)
Allen and Holbrook (2000)
Aspartateaminotransferase
Increased activity towardb-branched amino and 2-oxoacids and valine (>105-fold increase)
Yano et al. (1998),Oue et al. (1999)
Cre recombinase Altered specificity to recognize variants ofloxP recombination site (contribution to theunderstanding of protein-DNA recognition)
Santoro and Schultz (2002)
DNA polymerase Activity switched to RNA polymerase(efficient new enzyme activity)
Xia et al. (2002)
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267 259
the creation of hybrids (ITCHY), recombined extension on truncated templates (RETT),degenerate oligonucleotide gene shuffling (DOGS) and in vivo recombination.
The mutant genes obtained from directed evolution are transformed and expressed in asuitable host. Enzyme libraries are selected and/or screened for a range of selection properties(e.g., substrate range, stability in reactions or in nonaqueous solvents, thermostability). Genesobtained in the first round may be used as templates for the subsequent evolution cycles. Forsexual evolution, a pool of homologous genes (or genes generated using asexual methods) ispartially digested with DNAse-I and recombined by PCR. Alternatively, homologousrecombination can also be achieved in vivo based on the transformation of S. cerevisiaewith a linearized plasmid and target gene variants yielding a circular plasmid.
Whether mutations should be targeted to specific regions or distributed throughout thegene and methods of identifying improved variants (screening vs. selection) are the issues ofdebate. There is no agreement between different research groups as to a single best approachand there may never be one, since they all address different needs and situations. Manyreviews on various aspects of directed evolution citing various examples have appeared(Arnold, 2001; Benhar, 2001; Bornscheuer and Pohl, 2001; Brakmann, 2001; Brakmann andJohnsson, 2002; Jaeger et al., 2001; Powell et al., 2001; Reetz, 2001). Table 5 illustrates someexamples of important biocatalysts that have been improved using directed evolution. Severaldevelopments have been reported for the high-throughput screening of enzyme librariescreated by directed evolution (Baumann et al., 2001; Reetz, 2002b). Also, several algorithmsand computational models have been proposed to improve the efficiency of directed evolution(Bolon et al., 2002).
Most protein design efforts have focused on using only a single method. Rational design iswell suited to optimize direct interactions, but is not suitable for identifying distant mutations.Directed evolution, although optimizing activity, may not be suitable to introduce completelynovel activity. Therefore, it is beneficial to combine the powers of individual techniques.Some recent work has attempted to combine rational design and directed evolution toimprove biocatalysts (Bornscheuer and Pohl, 2001; Bolon et al., 2002).
Recently, directed evolution has been also applied for metabolic pathway engineering(Schmidt-Dannert, 2001; Zhao et al., 2002). The ability of directed evolution to simulta-neously transfer a large number of genes has been utilized to engineer new products. Thismethod forms the basis of the emerging area of ‘‘combinatorial biocatalysis,’’ which employsiterative reactions catalyzed by isolated enzymes or whole cells, in a natural or unnaturalenvironment, on substrates in solution or on a solid phase and harnesses the natural diversityof enzymatic reactions to generate libraries of organic compounds. Lipases and proteaseshave been employed in nonaqueous media to generate a library of acylated flavonoidderivatives and dibenzyl-1,2-phenylenedioxy diacetate derivatives (Rich et al., 2002).
5. Concluding remarks
Enzymatic reactions are no longer restricted to aqueous solutions. Chemists can now takeadvantage of enzyme specificities under mild conditions to catalyze reactions that were earlier
S. Hari Krishna / Biotechnology Advances 20 (2002) 239–267260
limited to using chemical catalysis. However, biocatalysis is still not viewed as a first-linealternative, but only as a last resort when other possible synthetic schemes fail. Nevertheless,there are numerous examples of industrial biocatalytic processes. Protein engineering isemerging as a major thrust area in improving enzyme activities and in finding novelapplications. A skillfully selected combination of chemical and biocatalysis is probably theway forward for many commercial syntheses. The global market for specialty enzymes wasabout US$1.5 billion in 1998 and is increasing with a predicted 5–10% growth per annum.Continuing improvements in biocatalysts and a better understanding of biocatalysis areexpected to greatly influence the production of fine chemicals.
Acknowledgements
The author is grateful to his mentors Prof. Dr. N.G. Karanth (Fermentation Technology andBioengineering, Central Food Technological Research Institute, Mysore, India) and Prof. Dr.Uwe T. Bornscheuer (Technische Chemie und Biotechnologie, Universitat Greifswald,Greifswald, Germany) for their encouragement and suggestions. The Alexander vonHumboldt Foundation (Bonn, Germany) is thanked for the award of Humboldt ResearchFellowship allowing the author to pursue research in Germany.
References
Ahern TJ, Klibanov AM. The mechanisms of irreversible enzyme inactivation at 100 !C. Science 1985;228:1280–4.
Akanuma S, Yamagishi A, Tanaka N, Oshima T. Serial increase in the thermal stability of 3-isopropylmalate dehydrogenase from
Bacillus subtilis by experimental evolution. Protein Sci 1998;7:698–705.
Allen SJ, Holbrook JJ. Production of an activated form of Bacillus stearothermophilus L-2-hydroxyacid dehydrogenase by
directed evolution. Protein Eng 2000;13:5–7.
Arnold FH. Protein design for non-aqueous solvents. Protein Eng 1988;2:21–5.
Arnold FH. Engineering enzymes for non-aqueous solvents. Trends Biotechnol 1990;8:244–9.
Arnold FH. Protein engineering for unusual environments. Curr Opin Biotechnol 1993a;4:450–5.
Arnold FH. Engineering proteins for non-natural environments. FASEB J 1993b;7:744–9.
Arnold FH. Combinatorial and computational challenges for biocatalyst design. Nature 2001;409:253–7.
Arnold FH, Moore JC. Optimizing industrial enzymes by directed evolution. Adv Biochem Eng Biotechnol 1997;58:2–14.
Asano Y. Overview of screening for new microbial catalysts and their uses in organic synthesis—selection and optimization of
biocatalysts. J Biotechnol 2002;94:65–72.
Baumann M, Sturmer R, Bornscheuer UT. A high-throughput-screening method for the identification of the active and enantio-
selective hydrolases. Angew Chem Int Ed Engl 2001;40:4201–4.
Bedell BA, Mozhaev VV, Clark DS, Dordick JS. Testing for diffusion limitations in salt-activated enzyme catalysts operating in